ThinhHoang1/claude-code-agent-team
An AI dev team for Claude Code agent teams: 8 role agents, a HANDOFF protocol and 3 hook quality gates — with the TeammateIdle infinite-loop bug fixed and regression-tested.
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code
$ npx degit ThinhHoang1/claude-code-agent-team/.claude ./rig-claude-code-agent-team # inspect, then merge into .claude/ MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ curl -fsSL --create-dirs -o .claude/agents/architect.md https://raw.githubusercontent.com/ThinhHoang1/claude-code-agent-team/main/.claude/agents/architect.md $ curl -fsSL --create-dirs -o .claude/agents/codebase-historian.md https://raw.githubusercontent.com/ThinhHoang1/claude-code-agent-team/main/.claude/agents/codebase-historian.md $ curl -fsSL --create-dirs -o .claude/agents/developer.md https://raw.githubusercontent.com/ThinhHoang1/claude-code-agent-team/main/.claude/agents/developer.md $ curl -fsSL --create-dirs -o .claude/agents/frontend.md https://raw.githubusercontent.com/ThinhHoang1/claude-code-agent-team/main/.claude/agents/frontend.md $ curl -fsSL --create-dirs -o .claude/agents/product-manager.md https://raw.githubusercontent.com/ThinhHoang1/claude-code-agent-team/main/.claude/agents/product-manager.md $ curl -fsSL --create-dirs -o .claude/agents/product-strategist.md https://raw.githubusercontent.com/ThinhHoang1/claude-code-agent-team/main/.claude/agents/product-strategist.md $ curl -fsSL --create-dirs -o .claude/agents/qa.md https://raw.githubusercontent.com/ThinhHoang1/claude-code-agent-team/main/.claude/agents/qa.md $ curl -fsSL --create-dirs -o .claude/agents/tech-lead.md https://raw.githubusercontent.com/ThinhHoang1/claude-code-agent-team/main/.claude/agents/tech-lead.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 7,204 rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(rm -rf *)",
"Read(**/*.pem)",
"Bash(rm -rf /)",
"Bash(git push --force:*)",
"Bash(sudo *)",
"Read(.env)",
"Bash(rm -rf /*)",
"Read(./.env.*)",
"Read(~/.aws/**)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(sudo:*)",
"Bash(git reset --hard*)",
"Bash(git reset --hard:*)",
"Read(.env.*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(rm:*)",
"Bash(git rebase *)",
"Bash(wget *)",
"Bash(npm publish:*)",
"Bash(git commit *)",
"Bash(gh pr merge *)"
]
}
} Subagents (8)
| architect model: inherit | Solution architect who judges how the system SHOULD be built. Reviews designs and implementations for soundness, finds architectural debt, chooses between technical options, and states how a design be |
| codebase-historian model: inherit | Expert on how the existing system ACTUALLY works today. Traces real code paths, finds where a change should hook in, explains why things are the way they are (git history), and predicts what breaks if |
| developer model: inherit | Software engineer who implements backend, library, and tooling changes with clean, tested code - services, data access, integrations, refactors, performance work. The hands on the keyboard for anythin |
| frontend model: inherit | Frontend engineer who builds and fixes user interface code - screens, components, client-side state, accessibility - and proves it works in a real browser. Use when a feature needs UI, when something |
| product-manager model: inherit | Product manager who owns scope and quality of outcome. Decides whether a feature is worth building, cuts scope, writes acceptance criteria as runnable user-level checks, and gives the final sign-off. |
| product-strategist model: inherit | Product strategist who holds the long-term direction and tests it against evidence. Answers "where is this product heading?", "does today's decision move us toward or away from that?", and "what evide |
| qa model: inherit | QA engineer who verifies quality by running the product like a real user - end-to-end flows, edge cases, failure paths - and gives a PASS / CONDITIONAL / FAIL verdict with evidence. Use after implemen |
| tech-lead model: inherit | Technical lead and orchestrator. Turns a vague request into an ordered plan, splits it into tasks with owners and definitions of done, assigns file ownership, spawns and coordinates the other roles, u |
Hooks (3)
| event | matcher | runs |
|---|---|---|
| TaskCreated | * | "$CLAUDE_PROJECT_DIR"/.claude/hooks/task-created-gate.sh |
| TaskCompleted | * | "$CLAUDE_PROJECT_DIR"/.claude/hooks/task-completed-gate.sh |
| TeammateIdle | * | "$CLAUDE_PROJECT_DIR"/.claude/hooks/teammate-idle-gate.sh |
Similar rigs
eugeek/claude-config-tdd
Claude Code config with TDD workflow, plan -> tests -> red -> implementation -> green -> review -> shipment
Orchestrator 5.8k tok ·
yeyo11/agentry
Run Claude Code as a service: REST API, web UI and multi-agent orchestration in one container
Pragmatist 5.2k tok ·
danielfarnose/claude-code-dev-loop
A multi-agent development loop for Claude Code: five specialized subagents plan, implement, verify and merge — with a QA gate that cannot edit the code it judges.
Orchestrator 1.2k tok ·
IntuitivePhella/AI-PhellOS
Portable AI development workflow template and CLI for spec-driven, test-aware software delivery with Codex, Claude Code, and other coding agents.
Orchestrator 3.1k tok ·